A construct for reading comprehension

According to Alderson (2000) the construct of what is involved in the reading process will have an important influence on my test design. This means the construct will determine which questions are being asked and how the answers are being interpreted. I decided to adopt categories of different question types or reading categories that have been used in the PISA study 2006 (Cresswell & Vayssettes, 2006). These have been developed to test for general reading comprehension and are being developed to account for functions of reading that are relevant for different situations in and outside of school. Nonetheless, I found these categories useful in the context of my study which I will present in some detail. The categories are the following:

36 1. Retrieving information

2. Developing an interpretation

3. Forming a broad general understanding

4. Reflecting on and evaluate the content of the text 5. Reflecting on and evaluating the structure of the text

The same skills and subskills might be involved in answering questions from these categories but they are still helpful to diagnose reading comprehension as most readers should be able to answer those of type of questions to some extent (Cresswell & Vayssettes, 2006). I will in the following describe the categories and elaborate on how I predict that multimodal representations can help to answer these kinds of questions.

2.4.5.1 Retrieving information

While reading in everyday life might require that we only need some specific information from a given text (Cresswell & Vayssettes, 2006), the literal information in literary texts is the basis for interpretation or making inferences. In order to understand a text and its tone, we need to have information about details that are for example describing the scenery or depicting the features of a character. To retrieve specific information the reader needs to be able to scan a text and the relevant information is often found in a single sentence (Cresswell & Vayssettes, 2006). I hypothesise that questions from this category can be answered more easily if the given text is supported by an illustration as some detail might be featured in the illustration so that even if the corresponding text passage proves to be difficult to understand, the reader can still extract information about these details. Readers who read might skip individual words or pay less attention to some especially those they do not know or are not able to decode successfully. An audio version does decode all the individual words indiscriminately and at least identifies their pronunciation. It will, therefore, be easier for readers to attend to the meaning of individual words and to move on to higher-level processes.

2.4.5.2 Developing an interpretation

In order to develop an interpretation, the reader needs to go beyond a first impression of the text and must organize individual pieces of information into a coherent model of the text. They might also have to decide on how pieces of information are connected with or without markers indicating these relations (Cresswell & Vayssettes, 2006). As developing an interpretation depends on the reader to have understood most of the literal message of the text, this category will be supported by multimodal text representations because they support the

37

reader in accessing the literal meaning of a given text. Moreover, a picture has other means of depicting an atmosphere or the relationship between characters. If we assume that the illustrator has made correct inferences about a given story, then the students can follow his interpretation and only need to decode information encrypted in a picture which might be easier than interpreting verbal clues. I asked for example about the mood in a certain scene. By the choice of color and facial expression, the illustrator had given the scene a positive atmosphere that mirrored what was being said in the text. The same goes for the audio file of the text. Someone reading the text will always interpret the text at hand and will, therefore, provide meaning which might not be as obvious in the written text especially when the text is not in the reader’s own language. Voices might, for example, convey emotions such as anger, fear or excitement.

2.4.5.3 Forming a broad general understanding

In order to form a broad general understanding, the reader must be able to recognize which ideas presented in the text are central and which are of minor importance. They can demonstrate this by for example choosing the main topic or theme of a text or elaborating on an author’s intention (Cresswell & Vayssettes, 2006). The illustrations in the texts I have chosen for the reading test have been designed to support comprehension. They, therefore, deal with the main idea and the aspects of the story that are most relevant. Attending to the pictures will, therefore, give the students an idea about central aspects of the story, and they will not have to make the decision themselves.

2.4.5.4 Reflecting and evaluating the content of a text

This category asks the readers to activate their prior knowledge or their background knowledge. Readers might be asked to evaluate the content and to compare it with what they know about the world or with other sources of information on this topic (Cresswell & Vayssettes, 2006). As I hypothesize that pictures or illustrations will help to activate content schemata, added illustrations will help the readers to connect their own knowledge of the world with a given text. A spoken version will give them an idea of the context of the story because different accents will make it easier to place a text in a cultural context and thus again activate content schemata.

2.4.5.5 Reflecting on and evaluating the form of a text

This category is closely connected to genre and register. The reader is asked to comment on the author's style, how the author accomplishes his or her goals and which attitudes are

38

expressed in a text (Cresswell & Vayssettes, 2006). Pictures can give important information about the genre of a text. The way characters are being portrayed provide an idea about the intentions and attitudes of the author as seen by the illustrator. This can be of course problematic if attitudes, for example, are ambiguous and the illustrator choses one that can be misleading. In the case of the texts I have chosen, the pictures make it quite clear what kind of stories we are dealing with. In style, the illustrations resemble old oil paintings thus indicating an old-fashioned novel which enables students to activate schemata on this genre and help them to place the novel in a context of other novels and stories of this genre.

39

3 Methods and Materials

This chapter deals with the participants of my study and a detailed description of the different aspects of the study and choices made which are linked to the design chosen to answer the questions raised in the conclusion of the theory part. Moreover, I will give a thorough account of the different sessions in which I carried out the study. The chapter will conclude with a short description of the methods for analysing the data gathered in the different tests.

As general reading proficiency has been argued to be of importance for the effect of the multimodal texts in the previous chapter, I needed to measure the proficiency level of the students independently of the reading comprehension test which composed the main part of my study, In order not to exhaust students more than necessary, the proficiency level was measured using a standard and well-established reading test (the Hodder test, see below) in a separate session about one week before the main reading comprehension test. . The test for general reading proficiency was an established reading test which is mostly used with students who have English as a first language while I developed the second reading test myself. In order to do so I used categories of reading which have also been used for the PISA tests (Cresswell & Vayssettes, 2006). I developed questions in different formats and pre-tested a first version with a group of students (50) which did not participate in the study. Both parts of the test will be described in more detail below.